Some tumors, most notably prostate cancer, are associated with new bone formation around metastatic tumor deposits. Previous evidence indicates that the mechanism by which this occurs is that these particular tumors produce growth regulatory factors which stimulate resident osteoblasts to make new bone. We have examined a number of human and rodent tumors for their capacity to stimulate new bone formation in vivo when they are implanted adjacent to bone surfaces in nude mice. We plan now to use one of these tumors to identify the mechanisms responsible for tumor-induced new bone formation. The tumor we plan to concentrate on initially is a rat adenocarcinoma which arises spontaneously in a strain of aging Lobund-Wistar rats from the male urogenital tract and which stimulates new bone formation when implanted adjacent to bone surfaces in Lobund- Wistar rats or nude mice. We plan to purify the growth factor(s) responsible for the new bone formation using in vitro assays. We then plan to clone and express the responsible gene and to determine the importance of the expressed protein by using two approaches. Firstly, we will use single cell cloning of PA-III cells to obtain low and high producers of the osteoblast stimulating activIty in vitro and confirm the effects in vivo by injecting these cells over the calvariae of nude mice. Secondly we will induce loss of function of the factor(s) in vitro by transfecting PA-III cells with antisense cDNA to the factor(s) and then determine the effects in vivo in nude mice. We also plan to determine the effects of the factor(s) in vivo by injecting it locally over the calvariae of normal mice or by transfecting the gene into Chinese hamster ovarian cells in vitro and then examining the local effects of over- expression of the factor(s) by the tumor cells when they are injected over the calvariae of nude mice. We have developed both of these models in our group. To determine the relevance of these findings to human cancers that typically stimulate new bone formation, we will screen fresh human prostate cancers for expression of these factors. Finally, we plan to inject the tumor cells into the left cardiac ventricle of nude mice using a technique we have described in the Project by Yoneda to determine if the tumor cells will spread to endosteal bone surfaces and cause the formation of osteoblastic metastases. Such studies may provide insights not only into mechanisms by which bone formation is regulated in these pathologic situations, but also important regulatory mechanisms responsible for normal bone formation.